Quantitative Image Quality Analysis of Soft Copy Displays

摘要

The purpose of this work was to develop a computer based method for measuring the image quality of video displays used in digital mammography. The significance of this ability is that it will enable ready optimization of display image quality, and since the display is the weak link in digital mammography, this is expected to lead to better mammographer performance in detecting breast cancer. The original proposal involved application of the Computer Analysis of Mammography Phantom Images (CAMPI) algorithm to digital images of mammography phantom displayed on the monitor and images by a CCD camera. We have improved significantly on this idea and implemented an ideal- observer based method which is more suitable for display evaluation than CAMPI. This method has been applied to investigate the image quality of two digital mammography monitors with different phosphor types. The effect of CRYT luminance and pixel contrast has been evaluated. The measurements are of remarkable precision1 allowing quantitative dependencies to be accurately demonstrated, no an extent never before possible. For example, the linear dependence of Signal- to-noise ratio on pixel contrast has been demonstrated to a high order or accuracy. We believe this method is novel and particularly advantageous for AMLCD displays that are being developed.